J. Med. Chem. 1995,38,3874-3883
3874
Inhibition of Monoamine Oxidase-B by SH-IndenoE1,2=cIpyridazines:Biological Activities, Quantitative Structure-Activity Relationships (QSARs) and 3D-QSARS Silvia Kneubuhler,t Ulrike Thull? Cosimo Altomare,$ Vincenco Carts,$ Patrick Gaillard,t Pierre-Alain Carrupt,t Angel0 Carotti,*,$ and Bernard Testa*pt Institut de Chimie Thkrapeutique, Section de Pharmacie, Universitl de Lausanne, CH-1015Lausanne-Dorigny, Switzerland, and Dipartimento Farmacochimico, Universita di Bari, Via Orabona 4, I-70126 Bari, Italy Received March 7, 1995@
A large series (66 compounds) of indeno[l,2-clpyridazin-5-ones (IPS)were synthesized and tested on their monoamine oxidase-A (MAO-A) and MAO-B inhibitory activity. All of the tested compounds acted preferentially on MAO-B displaying weak (nonmeasurable IC50 values) to high (submicromolar IC50 values) activities. The most active compound was p-CF3-3-phenylIP (IC50 = 90 nM). Multiple linear regression analysis of the substituted 3-phenyl-IPS yielded good statistical results (a2= 0.74; P = 0.86) and showed the importance of lipophilic, electronic, and steric properties of the substituents in determining inhibitory potency. Various comparative molecular field analysis studies were performed with different alignments and including the molecular lipophilicity potential. This led to a model including the steric, electrostatic and lipophilicity fields and having a good predictive value (q2= 0.75;r2 = 0.93).
Introduction Monoamine oxidase (MAO, EC 1.4.3.4.) is a flavin adenine dinucleotide (FADbcontaining flavoenzyme existing as two isozymes (MAO-A and MAO-B) and playing a key role in the regulation of various physiological amines. This enzyme is also a target for therapeutic agents, MAO-A inhibitors being used as antidepressants, while MAO-B inhibitors like selegiline coadministered with L-DOPA are of interest in the treatment of Parkinson’s disease. Irreversible, mechanism-based inactivators of MA0 are considered obsolete due to severe side effects such as hepatotoxicity and the well-known “cheese effect”. Based on a better understanding of the biochemical and catalytic properties of MAO, recent research efforts have focused on selective and reversible inhibitors of MAO-A or MAO-
Scheme 1. General Synthetic Method Used in This Work r
B.l-8
Pyridazine derivatives are currently an object of sustained interest due t o the vast range of their potential activities as chemotherapeutics, antithrombotics, cardiovascular agents, antisecretory and antiulcer agents, analgetic and antiinflammatory agents, and central nervous system (CNS) stimulants or depres~ants.~JO In a previous paper, we reported the synthesis and activity of a series of novel 5H-indeno[l,2-clpyridazin-5-ones (IPS) acting as competitive inhibitors of MA0 and mainly MAO-B.11-13 Such IPS have a structural analogy (the endocyclic N-N functionality) with 2-aryl-1,3,4-oxadiazinesand 5-aryl-1,3,4-oxadiazoles which have been described as selective and competitive MAO-B inhibitors.1°J4J5 In order to improve the MA0 inhibitory activity of IPS and gain an understanding of their structure-activity relationships, we designed, prepared, and tested a large congeneric series of 3- and 4-substituted indeno[l,2 100 (20%) C
b d b > 10 (13%) e > 50 (34%) b >50 (34%) C
> 1 (129)
e e
>50 (27%)
f
74.4 i 2.5 20.7 k 1.9 83.5 i. 7.8 21.0 i. 0.6 > 100 (21%) 9.5 i 0.3 7.7 i 0.4 5.3 f 0.2 30.6 f 0.3 16.1 i 0.7 22.4 0.4 1.5 i 0.1 > 5 (26%) 2.18 i 0.08 32.5 i 3.4
*
f
d
>25 (38%) b 50.5 i 4.4
>25 (239) >25 (259) b b
c
C
C
C
e
e
f
f
Or percent inhibition in parentheses at maximum solubility. No inhibition at maximum solubility:
100 pM, 50 ,uM, 25 pM, e 10
,uM, f 1pM.
weaker MAO-B inhibitory activity than the parent compound 50 (Table 4). While more structural variations are necessary to derive reliable S A R s accounting for the role of rings A and B of the IP nucleus in modulating MAO-B inhibitory activity, substitutents at the pyridazine ring appear, at the moment, more important in determining both activity and selectivity. 2D-QSAR Study of MAO-B Inhibition. A classical 2D-QSAR of the 24 3-phenyl-substituted congeners (series B) was carried out using partial least squares (PLS) analysis, cluster analysis (CA), and multilinear regression (MR). In a first step, each substituent was described by 11parameters, primarily physicochemical in nature. Classical Hammet (a, and a,) and SwainLupton ( F and R) constants were used as electronic parameters, and the Hansch constant (n)was used as a descriptor of lipophilicity, while bulkiness was parametrized by the van der Waals volume (Vw). Parameters were taken from standard compilation^.^^,^^ Indicator variable (lortho) was also used to take into account the detrimental effect observed for ortho substitution (lortho = 1for ortho-substituted derivatives and 0 for the others). The 'H-NMR chemical shifts of the aromatic protons in the pyridazinyl and phenyl rings were also used. Since the biologically active conformation of the phenyl ring is unknown, it proved difficult to differenti-
ate between the two ortho or the two meta positions in the phenyl ring. Thus we defined that ortho and meta substitution occurs in position 2' and 3' and then consistently considered the chemical shifts of protons 5' and 6'. A preliminary analysis showed that A d ~ (Ad y is the difference between the proton chemical shifts of the substituted and unsubstituted X-phenyl derivatives) was reasonably correlated with u (r2= 0.75) but that A d ~ g and A 6 ~ were 4 not correlated with electronic parameters due to their strong dependence on conformational effects. Due to this correlation with u, the chemical shift A d ~ jwas , not considered in the statistical analyses. Using the restricted set of 10 structural descriptors, a PLS analysis with a leave-one-out cross-validation procedure afforded a three-component model ( n = 24; q2 = 0.76; r2 = 0.88; s = 0.278; F = 49). A cluster analysis of the loadings matrix was then performed in order to identify proximity relations between the 24 compounds in the three-component space. The resulting dendrogram is presented in Figure 1, revealing four groups of compounds and isolated points. The first group (compounds 7,33,39-42,44-48,50,51) contains meta- and para-substituted 3-phenyl-IPS bearing lipophilic (n> 0) substituents (halogens, trifluoromethyl, methoxy), while the second group (compounds 53, 54, 57,59) contains derivatives with hydrophilic (n< 0) but strongly electron-withdrawing substituents (cyano, ni-
Journal
3D-QSARsof MAO-B Inhibition
of
Medicinal Chemistry, 1995, Vol. 38, No. 19 3877
Table 3. MA0 Inhibition Data, Physicochemical Parameters Used To Derive Classical QSARs, and CoMFA Predictions for (Series B) Substituted 3-Phenyl-5H-indeno[l,2-clpyridazin-5-ones
Ic50‘ @MI
f
compd
X
7 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
H 2’-CH3 2’-OH 3’-OH 4‘-OH 2’-OCH3 3‘-OCH3 4’-OCH3 3’,4’-(OCH3)2 3’-OCH3, @-OH 3‘-NHz 4‘-NHz 2’-F 3‘-F 4‘-F 3‘,4‘-F2 2‘-C1 3’-C1 4’-C1 3‘,5‘-C12 3’-Br 4’-Br 2‘-CF3 3‘-CF3 4‘-CF3 3’,5’-(CF3)2 3’-CN 4’-CN 2’-N02 3’-NO2 4‘-NO2 4’-COOH 4’-COCH3 4‘-NHCOCH3 4‘-OC4Hg
63
H
MAO-A d
f g g
>50 (28%) d d
f g
f
25.9 f 3.3 36.8 f 2.9 220 (17%)
f
30.6 i 1.6 g
>50 (40%) > 5 (17%) > 5 (30%) g g g
f
> 5 (24%) g g >10 (17%) e C
g
e C
> 10 (22%) > 10 (12%)
4 L N 3
CoMFA residuals
residuals
model C7
model F7
H5’
H6’
H4’
21.0 f 0.6 30.9 f 1.8 >l(ll%) 21 (18%) 9.9 f 1.7 >50 (21%) 1.6 i 0.2 3.22 f 0.04
4.68 4.51
-0.62 0.08
-0.12 -0.17
-0.26 -0.19
0 -0.15 -0.51
0 -0.62 -0.32
0 -0.28 0.26
5.00
0.18
-0.54 -0.42 -0.08 -0.51
-0.06 -0.16 -0.49 -0.03
-0.04 0.16 0.01 -0.05
-0.05 -0.22 -0.74 -0.31 0.04 -0.30 -0.19 -0.01 -0.04 -0.02
-0.58 -0.71 -0.13 0.05 -0.24 0.01 -0.29 -0.37 -0.13 -0.05 -0.10 -0.09 -0.12 -0.28 0.19 0.13 0.49 0.22 0.12 -0.34 0.40 0.31
-0.02 0.27 -0.07 0.09 -0.03 -0.03 -0.03 -0.01 -0.02 -0.02 -0.04 -0.02 -0.02 -0.28 0.04 0.04 0.08
0.12
0.07 -0.01 -0.06 -0.06
0
0.06
5.80 4.49
-0.07 -0.06
0.21 -0.01
0.24 -0.11
4.49 4.61 4.74 5.38 5.60 6.04 4.80 6.05 6.04 6.15 6.18 5.88
-0.03 0.24 -0.18 -0.23 0.09 0.18 0.18 0.11 -0.11 -0.42 0.15 -0.43
-0.04 -0.04 0.07 -0.06 0.09 -0.07 0.10 0.20 0.11 -0.02 0.17 -0.14
-0.12 0.04 0.21 -0.17 -0.02 -0.15 0.25 0.20 0.20 -0.05 0.20 -0.02
6.55 7.05
0.48 0.59
-0.09 -0.17
-0.08 0.03
g
f
32.3 f 4.1 24.7 f 2.4 18.1 f 1.5 4.2 f 0.2 0.92 i 0.06 0.92 5 0.06 15.9 f 1.0 0.90 f 0.08 0.92 i 0.10 0.71 i 0.02 0.66 i 0.05 1.33 2c 0.09
f
0.28 f 0.02 0.09 f 0.01 2 2 (15%) 3.7 5 0.3 2.2 f 0.2 80.1 i 4.4 >0.5 (42%) 0.50 f 0.02
5.43 5.66 4.10
0.07
0.13
-0.10
0
-0.18
-0.14
6.30
0.29
5.75
-0.20
-0.02
0.09 0.03 -0.36 0.15
1.78 f 0.09
0
-0.14
> 10 (9%)
f
f g
23.4 f 2.4
4.63
*
Table 4. MA0 Inhibition Data of Compounds 64-66 Structure
1.31i ~
0.08
5.10i ~
a
-0.12 0.14 0.06 0.14 0.28 0.14 0.31 0.17 0.22 0.79
0.01
0.03 -0.30 0.09 0.09
C
0.23
No inhibition at maximum solubility: 25 pM, 5 ,uM.
tro, acetyl). A third group can be observed which contains hydrophilic hydrogen-bond donor substituents (compounds 31,34,37,38) in the meta orpara position. The fourth group contains two ortho derivatives (com-
0.60 0.17 -0.51 -0.52
0 -0.05
-0.07
-0.07
Or percent of inhibition in parentheses at maximum solubility. No inhibition a t maximum solubility: 10 pM, g 1 pM. Different chemical shifts relative to the parent compound 7.
Cpd.
Ad
pICsO
g
(31%)
2D-QSAR
MAO-B
100 ,uM, 50 pM, e 25 pM,
pounds 28,43). The less active compound 55 (0-NOz) is separated from the previous clusters presumably because the electron-withdrawing capacity of the nitro substituent cannot compensate for the detrimental steric effect of ortho substitution. Although the PLS model is characterized by a good predictive capacity, its interpretation in physicochemical terms remains difficult. We therefore decided to perform a multilinear analysis with the most suitable parameters. The selection of the pertinent structural parameters was done by considering the similarity between parameters and their relative influence on the PLS model. A cluster analysis on the scores matrix of the PLS analysis given in Figure 2 indicates that some structural parameters contain comparable information, namely (T and F or Iortho and Vw2‘. Indeed the contribution of some parameters such as R and A h to the PLS model was very low, respectively 1%and 5%. Thus a stepwise multilinear regression was performed using only six parameters (0,n,A h , Vw2’, Vw3’, and Vw4‘).
3878 Journal of Medicinal Chemistry, 1995, Vol. 38, No. 19
37
Kneubuhler et al.
I 6+6+
31 59
Alignment 2
Figure 3. Rules for alignments 2 and 3. The torsion angle r
57 54 53 51 48 45
between the indenopyridazine ring and the phenyl ring in position 3 is fixed at 116".The difference between alignments 2 and 3 lies in the position of the meta substituent; for alignment 2 the meta substituent was considered in the 3' position, whereas for alignment 3 it was in the 5' position.
-
46 60
-
-
47
Iortho
Alignment 3
J
t I I I I I I I I I I I I I I I I / I I /
0.0
0.25
0.5
0.75
1.o
Figure 2. Cluster analysis of the score matrix of the 24 MAO-B inhibitors in series B, revealing which structural parameters contain comparable information, e.g., a and F or Iortho and Vw2'.
The stepwise regression analysis retained only four variables to produce eq 1 (95% confidence limits in parentheses):
+
PIC,, = 0.66(&0.36)~ 0.55(&0.23)~-
+
1.25(&0.40)Vw2' 0.31(f0.27)Vw4'
+ 5.51(&0.23) (1)
n = 24; q2 = 0.74; r2 = 0.86; s = 0.31; F = 29
The above results relate MAO-B inhibitory activity mainly t o lipophilic and electronic properties of substituents and to a steric negative effect of ortho substituents. A small but significative contribution is also noted for the volume of para substituents. An extension of the 2D-QSAR to the 39 compounds measured was not possible due to a different pattern of substitution on the 3 and 4 positions of the pyridazinyl ring. A 3D-QSAR study was therefore undertaken. 3D-QSAR Study of MAO-B Inhibition. Since its introduction in 1988,20comparative molecular field analysis (CoMFA) has proven to be a useful tool in 3DQSAR studies. In the first step of this study, we used the same compounds as in the above 2D-QSAR investigation (i.e., the 24 compounds of series B, see Table 3). This allowed us to derive a 3D-QSAR model which could readily be compared with eq 1. In a second step, the compounds in series A (Table 2) and series B were pooled together (i.e., 39 compounds with well-defined IC,, values) to gain more insights into the structural requirements modulating MAO-B inhibitory activity. The MAO-A data presented here were not suitable to a QSAR analysis due to the limited number of ICsos. Alignment. The superposition of all compounds is a critical step in any CoMFA study. As a first anchor point the very rigid indeno[l,2-clpyridazine rings were superimposed atom by atom. Then, we derived different alignments for several substituent orientations. No further alignment was necessary if the conformation of minimal energy was used, otherwise the substituents were aligned according to (a) the angle z between the indenopyridazine ring and the phenyl ring in position 3, (b)the topology of the substituents on the phenyl ring, and (c) the orientation of these substituents. For the torsion angle z between the indenopyridazine ring and the 3 substituent, a conformational analysis showed the existence of two zones (75"-140", 260"-320") of low energy within which the angle could be modified without significant difference in energy. Three alignments were retained (because of their maximal predictive power) among the many ones tested. For alignment 1,the indeno[l,2-clpyridazine rings were superimposed with their substituents in the minimumenergy conformation. The alignment rules for alignments 2 and 3, which take into account the unknown orientation of ortho and meta substituents, are presented in Figure 3. The torsion angle t between the indenopyridazine ring and the phenyl ring in position 3 was kept at the average minimum value of 116" calculated for all the ortho-substituted compounds. We also derived a model based on alignment 3 and defined by the field fit option, but its quality was low (results not shown).
3D-QSARsof MAO-B Inhibition
Journal of Medicinal Chemistry, 1995, Vol. 38, No. 19 3879
Table 5. Statistical Results of CoMFA Analyses of 24 Indeno[ 1,2-c]pyridazines (Series B)
model
q2
Nc
r2
relative contribution ste ele lip0
A3 A4 A5 A6 A7
alignment la steric field (ste) electrostatic field (ele) lipophilic field (lipo) ste and ele ste and lip0 ele and lip0 ele, ste and lip0
0.26 0.26 0.30 0.42 0.69 0.57 0.83
B1 B2 B3 B4 B5 B6 B7
alignment 2d ste ele lip0 ste and ele ste and lip0 ele and lip0 ele, ste, and lip0
0.27 0.40 0.20 0.48 0.64 0.78 0.84
0.77 0.44 0.56 0.87 0.65 0.35 4 0.94 0.62 0.38 4 0.99 0.37 0.35 0.28
alignment 3d ste ele lip0 ste and ele ste and lip0 ele and lip0 ele, ste, and lip0
0.23 0.30 0.35 0.50 0.64 0.63 0.84
2 3 3 2 3 4 3
A1
A2
c1 c2 c3 c4 c5 C6 c7
1 1
4 2 3
0.67 0.38 0.62 0.90 0.62 0.38 1 0.78 0.64 0.37 4 0.98 0.36 0.37 0.27
2 4 2 2 2
0.78 0.45 0.55 0.92 0.61 0.39 0.92 0.68 0.32 0.98 0.36 0.36 0.28
a The indenopyridazine rings were superimposed with their substituents in the minimum energy conformation. Crossvalidated correlation coefficient. Optimal number of principal components used for the final analysis. See Figure 3.
3D-QSAR of Series B (24 Compounds). The statistical results obtained with the three alignments are presented in Table 5. With each field separately, no good statistical results were obtained (q2i 0.4). In contrast, models incorporating two fields showed a significant improvement, with the lipophilicity field making a n important contribution (models A5, A6, B5, B6, C5, C6) relative to the combined electrostatic and steric fields (models A4, B4, C4). However, the best models were obtained by combining the three fields (electrostatic, steric, lipophilic) (A7, B7, C7), with q2 increasing by about 0.2. It should also be noted that the three alignments gave almost the same model with a identical predictive quality (q2= 0.83, 0.84, and 0.84 for models A7, B7 and C7, respectively). From a statistical viewpoint, model C7 is slightly better than the other two since it is based on a smaller number of components. Thus the different alignments influenced only slightly the statistical results but obviously had a large influence on the graphical results, perhaps suggesting a certain flexibility of the active site. Indeed the graphical representations of model C7 (Figure 4a) show that substitution in the meta or para position produces a sterically favorable (green) contribution, whereas substitution in the ortho position is sterically unfavorable (red) probably due to steric hindrance. The electrostatic contributions consist mainly (a) in an important magenta zone near the meta and para positions of the 3-phenyl ring where high electron density is favorable to activity and (b) a n important white zone near the bond between the aromatic carbon and the para substituent where electron deficiency is favorable to activity (Figure 4a). The statistical results of the lipophilic field reveal only lipophilic (yellow)zones favorable for MAO-B inhibitory activity near the meta and para substituents.
c
I
Figure 4. Pictorial representation of 3D-QSAR models. The color code is as follows: sterically favorable and unfavorable influences, green and red zones, respectively; favorable influence of high electron density and deficiency, magenta and white zones, respectively; favorable influence of lipophilic and hydrophilic (polar) groups, yellow and cyan zones, respectively. (a) Model C7 for the 24 MAO-B inhibitors of series B. Substitution in the meta and/or para position has a favorable steric contribution, whereas substitution in the ortho position is sterically unfavorable. The important magenta zone near the meta and para positions indicates t h a t a high electron density is favorable for activity, whereas electron deficiency is favorable between the aromatic carbon atom in the 4' position and the para substituent. Moreover lipophilic substituents in the meta and para positions are favorable for activity. Contour levels: steric field, 20% red, 75% green; electrostatic field, 20% magenta, 85% white; lipophilic field, 0% cyan, 85%yellow. (b) Model F7 for the 39 MAO-B inhibitors in series A plus B. A bulky substituent in the para and/or meta position can make a favorable contribution, whereas ortho substitution is unfavorable. Electron-withdrawing substituents in the para a n d o r meta position produce favorable electronrich and electron-poor zones. Hydrophilic substituents in position 4 appear to be favorable for activity. Contour levels: steric field, 20% red, 80% green; electrostatic field, 15% magenta, 85% white; lipophilic field, 5% cyan, 85%yellow.
As explained in the Experimental Section, compounds 16 and 56 were taken as the test set. Using model C7, the IC50 predictions for compound 16 (PICSoestim = 4.851 PICSOpred = 4.87) and compound 56 (pIC5oestim = 6.16/ PICSOpred = 5.88) were quite satisfactory. Comparison of Q S A R and CoMFA Models for Series B (24 Compounds). This comparison leads t o
3880 Journal of Medicinal Chemistry, 1995, Vol. 38, No. 19 Table 6. Statistical Results of CoMFA Analyses of 39 Indeno[l,2-clpvridazines(series A and B) relative contribution model
D1 D2 D3 D4 D5 D6 D7
El E2 E3 E4 E5 E6 E7
F1 F2 F3 F4 F5 F6 F7 a-d
Ne
r2
0.49 0.37 0.50 0.59 0.67 0.65 0.78
3 6 1 2 2 1 3
0.72
alignment 2d ste ele lip0 ele and ste ste and lip0 ele and lip0 ele, ste, and lip0
0.45 0.35 0.43 0.55 0.71 0.49 0.70
7 2 3 2 3 4 3
alignment 3d ste ele lip0 ele and ste ste and lip0 ele and lip0 ele, ste, and lip0
0.39 0.36 0.46 0.60 0.55 0.65 0.75
2 3 1 3 2 2 3
q2
alignment 1" steric field (ste) electrostatic field (ele) lipophilic field (lipo) ele and ste ste and lip0 ele and lip0 ele, ste, and lip0
ste
ele
lip0
0.61 0.78 0.48 0.52 0.82 0.56 0.44 0.78 0.53 0.47 0.94 0.28 0.37 0.35 0.86 0.77 0.76 0.40 0.60 0.92 0.46 0.54 0.86 0.54 0.46 0.91 0.22 0.43 0.35
0.57 0.84 0.46 0.54 0.75 0.57 0.43 0.87 0.49 0.51 0.93 0.27 0.41 0.32
See Table 5 .
the following conclusions. (1)Ortho effect: The detrimental effect of ortho substitution found in the classical QSAR study and expressed by the structural parameter Vw2' is clearly attributed in the CoMFA study to a corresponding region of steric hindrance. (2) Steric and lipophilicity parameters: The favorable influence on activity noted in the QSAR study for the hydrophobic constant n and the volume of para substituents is also found in the CoMFA models. Indeed, increased activity is associated with a favorable steric zone (green in the figure) near the para position and with a favorable liphopilic zone (yellow). (3) Electronic parameters: The presence of a white region near the bond between the 3-phenyl ring and its para substituent is compatible with the favorable influence of u identified in the 2D-QSAR model. The greater the value of o for apara substituent, the higher the positive atomic charge on the para aromatic carbon (C4'). Thus the increase of this atomic charge is reflected by a positive (white) zone in the CoMFA models. The magenta zone is probably associated with an additional anchor point within the MAO-B active site. However, its exact nature (electrostatic interaction or hydrogen bond) is not yet established. 3D-QSAR Study with the Combined Series A and B. Compiling series A and B (39 compounds with useable activities), three models based on the alignments described above, were derived. As reported in Table 6, a single field alone yields a satisfactory q2 in a number of cases (models D1, D3, E l , E3, F3). Again the contribution of the lipophilicity field proved important in all models with two fields (models D5, D6, E5, E6, F5, F6). The best models were obtained with alignments 1or 3 using the three fields (models D7, F7), with model D7 having the best predictive quality of all (q2= 0.78). However, model F7 which is almost as good was chosen as the final one since the orientation of ortho
Kneubiihler et al.
and meta of minimized compounds (alignment 1) was almost random, which renders the results of D7 less clear. The graphical representation of model F7 (Figure 4b) shows that all the favorable and unfavorable zones of the three fields occupy well-distinct sectors in space and are therefore not intercorrelated. As with model C7, a bulky lipophilic substituent in the para or meta position of the 3-phenyl ring can make a favorable steric contribution (green zone), whereas a n ortho substituent decreases MAO-B inhibitory activity (red zone). Electronwithdrawing substituents in para and meta positions are doubly favorable for activity, since they induce a favorable region of high electron density close to the substituents themselves (magenta zone) and a favorable region of electron deficiency close to the 4'-carbon (white zone). A region of favorable hydrophilic contribution is seen above and below ring C of the indenopyridazine system (cyan zone) presumably expressing the influence of substituents in the 4 position. Indeed, lipophilic 4-substituents decrease MAO-B inhibitory activity, whereas polar 4-substituents increase it. Model F7 is quite satisfactory in predicting the IC50 of compound 16 (PIC~~estim = 4.85/pIC5opred= 4.69) and compound 56 (pIC5oestim = 6.16/PIC50pred= 5.80). The statistical results and the combined steric, electrostatic, and lipophilic contributions in models D7, E7, and F7 are in agreement with the CoMFA study of series B. Indeed, models D7, E7, and F7 do not contain important additional information relative to models A7, B7, and C7 for the 3-phenyl ring and its substituents, but they alone explore variations around ring C of the indenopyridazine ring. In contrast, this study did not generate any information regarding substitution of rings A and B.
Conclusion 5H-Indeno[l,2-clpyridazin-5-ones are shown here to be competitive, reversible MA0 inhibitors with a relatively high selectivity for MAO-B. The most potent compound (51) is comparable to some of the best inhibitors known to date, e.g., lazabemide (IC50 = 30 nMI2l and some time-dependent inhibitors of the oxadiazo1 series (where the most active compound has an IC,, of 1.4 nM after 60 min).15 Congruent 2D- and 3DQSAR analyses allowed to gain insight into the key structural features governing MAO-B inhibitory activity of indenopyridazines. Experimental Section Chemistry. Melting points were determined by the capillary method on a Gallekamp MFB 595 OlOM electrothermal apparatus and are uncorrected. Elemental analyses were performed on a Carlo Erba 1106 analyzer; for C, H, and N the results agreed to within &0.40% of the theoretical values. IR spectra were recorded using potassium bromide disks on a Perkin-Elmer 283 spectrophotometer. Only the most significant IR absorption bands are listed. 'H-NMR spectra were recorded at 300 MHz on a Bruker 300 instrument for the indeno[l,2-c]pyridazin-5-one derivatives and at 90 MHz on a 390 Varian spectrometer for intermediate aldol adducts (Scheme 1). Chemical shifts are expressed in d (ppm) and the coupling constants J in hertz (Hz). The following abbreviations are used, s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; dd, double doublet; dt, double triplet; td, triple doublet; qd, quadruple doublet; tdd, triple double doublet; br, broad signal. Signals due to OH and NH protons were located by deuterium exchange with DzO.
3D-QSARsof MAO-B Inhibition
Journal of Medicinal Chemistry, 1995, Vol. 38, No. 19 3881
IR and 'H-NMR spectra were fully consistent with the proposed structures. The synthesis of compounds 1, 4,7,12, 25,34,57; 8,9,15,18,19,21,23,24,28-30, 41,43-45, 47, 50, 56,58; and 2,3,5,6,10,14,16,17,20,31,39,53,55 has been described by Kneubiihler et al.I7 and Carotti et al." Synthesis of BH-Indeno[1,2-clpyridazin-5-ones 33,36,
3-[3'-(Trifluoromethy)phenyll-7-methoxy-5H-indeno[1,2-c]pyridazin-5-one (65):method A yield 60%; mp 214-5 "C, from ethanol; 'H-NMR (CDC13) d 8.40 (s, l H ) , 8.30 (d, l H , J = 7.7), 7.98 ( s , l H ) , 7.81 (d, l H , J = 8.41, 7.78 (d, l H , J = 7.7),7.69(t,lH,J=7.7),7.67(d,lH,J=2.3),7.03(dd,lH, J = 8.4, 2.3), 4.01 ( s , 3Hj; IR (cm-I) 1720, 1405, 1330. Anal.
40,42,46,49,51,52, and 65. General One-PotProcedure (MethodsA and B). A solution of ninhydrin (1.78 g, 10 mmol) and a suitable ketone (10 mmol) in 50 mL of glacial AcOH was heated to reflux until a check by TLC showed the
(C19HllF3N202)C, H, N. The 8-methoxy positional isomer was detected by HPLC-MS (5%).
complete disappearance of the starting materials (normally 1-3 h). The reaction mixture was allowed to cool to room temperature, and then 0.74 mL (15 mmol) of hydrazine hydrate (98%) was slowly added under magnetic stirring (method A). Such an addition could also be done after a 140-4 dilution with acetonitrile (method B). The completion of the reaction was normally obtained within 2-4 h. Title compounds were separated as solids from the reaction mixture, either directly or, if necessary, upon addition of few drops of water.
Synthesisof Aldol Adducts 22a,35a,48a,54a,59a,60a, 61a,and 62a. The preparation of the title compounds followed the described one-pot procedure up to the disappearance of
starting reagents ninhydrin and ketones (TLC check). The workup method was as follows. (ij Adduct 22a: the solid product, which precipitated from the reaction mixture upon cooling, was collected by filtration and recrystallized. (ii) Adducts 35a,48a,54a,59a,60a,61a,and 62a:the reaction mixture was evaporated to dryness under vacuum, and the solid residue was purified by crystallization (35a,48a,54a, 61a,62a)or flash chromatography on silica gel using CHC13 (59a,60a)as eluent. Reaction yields and physicochemical and 3-(3'-Methoxyphenyl)-5~-indeno[l,2-clpyridazin-5-spectroscopic data of the isolated aldol adducts are listed in Table 6. one (33):method B;yield 41%; mp 159-60 "C, from ethanol; Synthesis of 5H-Indeno[1,2-c]pyridazin-5-ones 22,35, 'H-NMR(CDC13)68.19(qd,lH,J=7.5,1.0,0.8),7.99(~, lH), 48, 54,59, 60,61,and 62 (MethodC). Hydrazine hydrate 7.87(qd,lH,J=7.5,1.2,0.8),7.78(dd,lH,J=2.6,2.0),7.75 98% (0.50 mL, 10 mmol) was added under magnetic stirring (td, l H , J = 7.5, 1.2), 7.63 (qd, l H , J = 8.0, 2.0, 0.91, 7.57 (td, to the solution of an aldol adduct (10 mmol) in anhydrous l H , J = 7.5, l . O ) , 7.45 (t, l H , J = 8.01, 7.08 (qd, l H , J = 8.0, EtOH (100-200 mL) containing 0.1 mL of CF3COOH. The 2.6, 0.91, 3.91 (s, 3Hj; IR (cm-l) 1730, 1415. Anal. (Cp,H12N202) yellow solid which was formed within 1-3 h was collected by C, H, N. 3-(3'-Methoxy-4-hydroxyphenyl)-5H-indeno[l,2-clpy- filtration and recrystallized. 3-(2'-Carboxyethyl)-5H-indeno[ 1,2-clpyridazin-5-one ridazin-&one (36): method B;yield 24%; mp 253-4 "C dec, (22):yield 72%; m p 200-1 "C dec, from ethanol; 'H-NMR from ethanol; 'H-NMR (CDC13j d 8.16 (dd, l H , J = 7.7, l . O j , (acetone-&) d 12.25 (br, l H , D20 exchange), 8.05 (dd, l H , J = 7.96 ( s , l H ) , 7.92 (d, l H , J = 2.01, 7.85 (dd, l H , J = 7.7, l.O), 8.3, l . O ) , 7.82 (dd, l H , J = 8.3, l.O), 7.80 (td, l H , J = 8.3, J = 7.74 (td, l H , J = 7.7, 1.01, 7.55 (td, l H , J = 7.7, 1.01, 7.54 (dd, LO), 7.78 (s, lHj, 7.63 (td, l H , J = 8.3, J = LO), 3.21 (t, 2H, l H , J = 8.3, 2.0), 7.48 (d, l H , J = 8.31, 5.91 (s, l H , DzO J = 7.5), 2.80 (t, 2H, J = 7.5); IR (cm-l) 3020 (brj, 1730, 1710, exchange), 4.01 ( s , 3H); IR (cm-') 3180 (br), 1725, 1415. Anal. 1425, 1415. Anal. ( C I ~ H I O N ZC, O ~H, ) N. (Ci8HizNz03) C, H, N.
3-(3'-Fluorophenyl)-5H-indeno[ 1,2-clpyridazin-5-one (40):method A; yield 64%; mp 215-6 "C, from ethanoudioxane (95:5); 'H-NMR (CDC13) 6 8.17 (dd, l H , J = 7.1, 1.01, 7.95 ( s , l H ) , 7.88 (qd, l H , J = 8.3, 2.5, 1.81,7.91-7.81 (m, 2H), 7.73 (td, l H , J = 7.1, l.O), 7.56 (td, l H , J = 7.4, 1.01, 7.49 (td, l H , J = 8.3, 6.2), 7.21 (tdd, l H , J = 8.3, 2.5, 1.0); IR (cm-lj 1720, 1430, 1410. Anal. (C17HgFN20) C, H, N.
3-(3',4'-Difluorophenyl)-5H-indeno[ 1,2-clpyridazin-5one (42): method A; yield 63%; mp 260-1 "C, from ethanol/ dioxane (9O:lO); 'H-NMR (CDC13) d 8.19 (dd, l H , J = 7.6, 0.91, 8.05 (qd, l H , J = 11.2, 7.5, 2.21, 7.95 (s, lH), 7.87 (dd, l H , J = 7.6, 0.9), 7.84 (m, l H ) , 7.76 (td, l H , J = 7.6, 0.91, 7.58 (td, l H , J = 7.6, 0.9j, 7.34 (dt, l H , J = 9.8, 8.5); IR (cm-') 1730,
3-(3',4-Dimethoxyphenyl)-5H-indeno[1,2-clpyridazin&one (35):yield 82%; mp 257-8 "C, from THF/ethanol (50: 50); 'H-NMR (DMSO-&) 6 8.42 (s, l H ) , 8.11 (dt, l H , J = 7.4, LOj, 7.92-7.80 (m, 4Hj, 7.66 (td, l H , J = 7.4, 1.01, 7.12 (d, l H , J = 8.41, 3.88 (s, 3Hj, 3.84 (s, 3H); IR (cm-') 1720, 1420. Anal. (C19H14N203) C, H, N. 3-(4'-Bromophenyl)-5H-indeno[ 1,2-clpyridazin-5-one (48):yield 90%; mp 261-2 "C, from ethanol; IH-NMR (CDC13) 6 8.18 (dt, l H , J = 7.5, 1.0, 1.01, 8.00 (m, 2H), 7.96 ( s , l H ) , 7.86(qd,lH,J=7.5,1.2,1.0),7.75(td,lH,J=7.5,1.2),7.67 (m, 2H), 7.56 (td, l H , J = 7.5, 1.0);IR (cm-'1 1725, 1415. Anal. (C17HgBrN20) C, H, N.
3-(4'-Cyanophenyl)-5H-indeno[ 1,2-clpyridazin-5-one
(54):yield 93%; mp 298-9 "C, from ethanol; 'H-NMR (CDC13) 1430, 1415. Anal. (C17H8F2N20)C, H, N. 3-(3',5'-Dichlorophenyl)-5H-indeno[l,2-clpyridazin-5- d 8.24 (m, 2H), 8.20 (dt, l H , J = 7.5, 1.11, 8.01 (s, l H ) , 7.88 (dt, l H , J = 7.5, l . l ) , 7.84(m, 2Hj, 7.77(td, 1 H , J = 7.5, 1.11, one (46):method A; yield 76%; mp 291-2 "C, from ethanol; 7.59 (td, l H , J = 7.5, 1.1);IR (cm-') 2220, 1720, 1415. Anal. 'H-NMR (CDC13) d 8.21 (qd, l H , J = 7.5, 1.1, 0.81, 8.02 (d, (CiaHgN30) c, H, N. 2H, J = 1,9j, 7.94 ( s , l H ) , 7.87 (qd, 1 H , J = 7.5, 1.2, 0.81, 7.76 3-(4-Acetylphenyl)-5H-indeno[ 1,2-clpyridazin-5-one (td, l H , J = 7.5, 1.2), 7.58 (td, l H , J = 7.5, 1.11,7.51 (t, l H , J (59):yield 95%; mp 256-7 "C, from ethanol; 'H-NMR (CDC13) = 1.9); IR (cm-') 1705, 1400. Anal. (C17HsC12N20) C, H, N. d 8.24 (m, 2Hj, 8.21 (m, l H ) , 8.13 (m, 2H), 8.05 ( s , lHj, 7.88 3-[2'-(Trifluoromethyl)phenyll-5H-indeno[ 1,2-clpy(m, l H j , 7.77 (td, l H , J = 7.6, 1.2), 7.59 (td, l H , J = 7.6, 1.21, ridazin-&one (49):method B;yield 23%; mp 180-1 "C, from 2.68 (s, 3H); IR (cm-') 1730, 1670, 1415. Anal. (C19H12N2021 ethanol; 'H-NMR (CDC13j 6 8.22 (dd, l H , J = 7.7, 1.01, 7.87 C, H, N. (dd,lH,J=7.7,1.0j,7.84(dd,lH,J=8.4,1.1),7.76(td,lH, 3-(4'-Acetamidophenyl)-5H-indeno[ 1,2-clpyridazin-5J = 7.7, l . O ) , 7.70 (s, l H ) , 7.67 (dd, l H , J = 6.5, 1.41, 7.63(60): yield 76%; mp 338-9 "C, from THF; 'H-NMR 7.55 (m, 3Hj; IR (cm-') 1710,1410,1310. Anal. ( C I ~ H ~ F ~ N ~ Oone ) (CDC13j d 10.25 (br, l H , DzO exchange), 8.18 (dt, l H , J = 7.6, C, H, N. l , O j , 8.12 (m, 2H), 7.97 ( s , l H ) , 7.85 (dt, l H , J = 7.6, 1.01, 7.74 3-[4'-(Trifluoromethyl)phenyll-5H-indeno[ 1,2-c1py(td, l H , J = 7.6, J = 1.01, 7.70 (m, 2H), 7.55 (td, l H , J = 7.6, ridazin-&one (51):method A; yield 79%; mp 270-1 "C, from l . O ) , 2.13 (s, 3H); IR (cm-') 3380, 1720, 1690, 1595, 1415. Anal. ethanovdioxane (9O:lO); lH-NMR (CDC13) d 8.25 (d, 2H, J = (CigHi3N302) C, H, N. 8.1j, 8.21 (dt, l H , J = 7.5, LO), 8.02 (s, l H ) , 7.88 (dt, l H , J = 3-(4'-Butoxyphenyl)-5H-indeno[ 1,2-clpyridazin-5-one 7.5,1.0),7.81(d,2H,J=8.1j,7.76(td,lH,J=7.5,1.0),7.59 (61):yield 72%; mp 178-9 "C, from ethanol; 'H-NMR (CDC13) (td, l H , J = 7.5, J = 1.0); IR (cm-l) 1725, 1420, 1330. Anal. d 8.15 (dt, l H , J = 7.5, l.l),8.07 (m, 2H), 7.92 ( s , l H ) , 7.83 (Ci8HgF3NzO) C, H, N. 3-[3',5'-Bis(trifluoromethyl~phenyll-5H-indeno[l,2-c]- (dt, l H , J = 7.5, l.l),7.72 (td, l H , J = 7.5, 1.11, 7.53 (td, l H , J=7.5,1.1),7.02(m,2H),4.03(t,2H,J=6.5),1.79(m,2H), pyridazin-5-one(52):method B; yield 39%; mp 214-5 "C, 1.49 (m, 2H), 0.98 (t, 3H, J = 6.5); IR (cm-') 1730, 1410. Anal. from ethanol; 'H-NMR (CDC13) 6 8.61 ( s , 2H), 8.22 (dd, l H , J (CziHisNzOz) C, H, N. = 7.7, 0.8), 8.06 ( s , l H ) , 8.03 ( s , l H ) , 7.90 (dd, l H , J = 7.7, 3-(4-Piperidinylphenyl)-5H-indeno[ 1,2-clpyridazin-50.9), 7.79 (td, l H , J = 7.7, 0.9), 7.61 (td, l H , J = 7.7, 0.8); IR one (62):yield 20%; mp 224 - 5 "C, from ethanol; 'H-NMR (cm-l) 1720, 1390, 1280. Anal. ( C I ~ H ~ F ~ N C,~H, O N. )
3882 Journal of Medicinal Chemistry, 1995, Vol. 38, No.19
Kneubiihler et al.
(CDC13) 6 8.15 (dt, l H , J = 7.6, 1.11,8.05 (m, 2H), 7.92 ( s ,
on silica gel using ethyl acetatehexane (50:50) as eluent: yield
61%; mp 237-8 "C, from ethanol; 'H-NMR (CDC13) 6 8.18 (m, lH),7.83(qd,lH,J=7.6,1.2,1.1),7.72(td,lH,J=7.6,1.2), l H ) , 7.95 ( s , l H ) , 7.85 (m, l H ) , 7.73 (td, l H , J = 7.2, 1.01, 7.55 7.52 (td, l H , J = 7.6, l.l),7.01 (m, 2H), 3.33 (m, 4H), 1.69 (m, (m, l H ) , 7.54 (m, l H ) , 7.41 (dt, l H , J = 7.7, l.l),7.31 (t, l H , 6H); IR (cm-') 1720, 1410. Anal. (C22H19N30) C, H, N. 3-Phenyl-4-(ethoxycarbonyl)-5H-indeno[l,2~lpyridazin- J = 7.71, 6.83 (qd, l H , J = 7.7, 2.7, 1.11, 3.87 (br, 2H, D2O exchange); IR (cm-l) 3305, 3210, 1730 1635, 1415. Anal. 5-one(11). To a solution of ninhydrin (1.78 g, 10 mmol) in dimethoxyethane (25 mL) on molecular sieves (3 A), kept a t room temperature over 1 day, was added an equimolar amount of ethyl benzoylacetate. The resulting mixture was left to stand a t room temperature for 3 days and then filtered and diluted to ca. 0.025 M concentration with anhydrous ethanol (375 mL). An equimolar amount of hydrazine monohydrate was added, under magnetic stirring, and after 3 h , the solution obtained was evaporated to dryness under vacuum and the residue purified by flash chromatography using hexaneiethyl acetate (80:20) a s eluent: yield 20%; mp 113-4 "C, from ethanol; 'H-NMR (90 MHz, CDC13) 6 8.23 (d, l H , J = 71, 7.937.33 (m, 8H), 4.41 (9, 2H, J = 71, 1.27 (t, 3H, J = 7); IR (cm-') 1740, 1730, 1395. Anal. (C20H14N203)C, H, N. 3-Benzoyl-5H-indeno[ 1,2-c]pyridazin-5-one (13). 3-Benzyl-5H-indeno[l,2-c]pyridazin-5-one (HCA) (272 mg, 1 mmol) in acetic acid (2 mL) was refluxed with 130 mg (1.3 mmol) of chromium(VI) oxide for 2 h. After cooling the mixture was diluted with water (8 mL), extracted with CHC13, and evaporated under vacuum. The residue was purified by flash chromatography using CHC13 as eluent: yield 94%; mp 168-9 "C dec, from ethanol; lH-NMR (CDC13) 6 8.27 ( s , l H ) , 8.27-
(Ci7HiiN30) C, H, N.
3-(4-Aminophenyl)-5H-indeno[ 1,2-c]pyridazin-5-one (38). Compound 38 was synthesized as the isomeric compound 37 from 3-[4'-(N-carbobenzoxyamino)phenyll-5H-indeno[l,2clpyridazin-5-one. The red solid obtained was purified by crystallization: yield 58%; mp 295-6 "C, from butanol; IHNMR (CDC13) 6 8.15 (d, l H , J = 7.71, 7.98 (m, 2H), 7.91 ( s , l H ) , 7.83 (d, l H , J = 7.71, 7.72 (t, l H , J = 7.71, 7.53 (t, l H , J = 7.71, 6.79 (m, 2H), 3.99 (br, 2H, D2O exchange); IR (cm-') 3410,3390,1730,1630 1610,1425. Anal. (C17HllN30)C, H, N.
3-(Trifluoromethyl)-5H-indeno[ 1,2-c]pyridazin-5-one (64). Ethyl 4,4,44rifluoroacetoacetate (0.24 mL, 2 mmol) and
356 mg (2 mmol) of ninhydrin were refluxed in acetic acid (4 mL). After 1.5 h the solution was diluted with acetonitrile (12 mL) and 0.15 mL (3 mmol) of hydrazine monohydrate was added. The yellow solid formed was filtered and dissolved in 4 mL of trifluoroacetic acid. After stirring a t room temperature for 4 h, the solution was evaporated under vacuum, and the solid residue was purified by crystallization: yield 20%; mp 157-8 "C, from CHClhexane; 'H-NMR (CDC13) 6 8.24 (m, lH), 7.92 (m, l H ) , 7.89 (s, l H ) , 7.79 (td, l H , J = 7.0, 1.01, 8.21(m,3H),7.91(td,lH,J=7.5,1.1),7.79(td,lH,J=7.5, 7.64 (td, l H , J = 7.0, 1.0);IR (cm-') 1720, 1360, 1290. Anal. L l ) , 7.80-7.60 (m, 2H), 7.56-7.48 (m, 2H); IR (cm-') 1720, 1660, 1410. Anal. (ClsHloN202)C, H, N. (CI~H~F~N C,~H, O )N. 3-[3'-(Trifluoromethyl)phenyl]-7-hydroxy-5H-indeno13H-Benz[h]indeno[l,2-clcinnolin-13-one (26). 11,12[1,2-clpyridazin-5-one (66). 65 (712 mg, 2 mmol) was Dihydro-13H-benz[h]indeno[1,2-c]cinnolin-13-one (HCA) (284 refluxed in 10 mL of aqueous HBr (47%)/acetic acid (50:50) mg, 1mmol) in dioxane (5 mL) was refluxed with 227 mg (1.3 for 2 h. The mixture was filtered and extracted with CHC13. mmol) of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone over 1.5 h. The organic layer was evaporated under vacuum, and the After cooling the solvent was evaporated, and the residue was residue was purified by flash chromatography on silica gel purified by flash chromatography on silica gel using CHCld using CHCldmethanol(95:5) a s eluent: yield 25%; mp 351-2 ethyl acetate (9O:lO) as eluent: yield 90%; mp 261-2 "C, from "C, from ethanol; 'H-NMR (DMSO-&) 6 11.34 (br, l H , DzO ethanol; 'H-NMR (CDC13)6 9.38 (m, l H ) , 8.41 (d, l H , J = 9.01, exchange), 8.59 ( s , l H ) , 8.57 (d, l H , J = 8.0),8.50 ( s , l H ) , 7.92 8.14(dt,lH,J=7.3,1.2),7.95(d,lH,J=9.0),7.86-7.67(m, (d, l H , J = 8.01,7.81 (t, l H , J = 8.01, 7.77 (d, l H , J = 8.51, 4H), 7.65 (td, l H , J = 7.5, 1.21, 7.45 (td, l H , J = 7.5, 1.1);IR 7.44 (d, l H , J = 2.21, 6.99 (dd, l H , J = 8.5, 2.2); IR (cm-l) (cm-l) 1720, 1400. Anal. (C19HloNzO) C, H, N. 3200 (br), 1705, 1475, 1340. Anal. (C18HgF3N202) C, H, N. llH-Indeno[1,2-c]cinnolin-ll-one (27). 1,2,3,4-TetrahyPreparation of Rat Brain Mitochondria. Rat brain dro-11H-indeno[l,2-c]cinnolin-ll-one11 (236 mg, 1 mmol) in mitochondria were isolated according to the method of Clark22 dioxane (10 mL) was refluxed with 522 mg (2.3 mmol) of 2,3modified by Walther e t al.23 For further details, see Thull et dichloro-5,6-dicyano-1,4-benzoquinone for 2.5 h. After cooling, al.24 The protein content of the washed mitochondrial fraction the mixture was concentrated, and the residue was purified was determined according t o the procedure of L0w1-y~~ with by flash chromatography on silica gel using CHCldethyl bovine serum albumin a s a standard. acetate (9O:lO) a s eluent: yield 12%; mp 285-6 "C dec, from MA0 Inhibition. The method of Weissbach26was modified ethanol; lH-NMR (CDC13) 6 8.65 (m, l H ) , 8,50 (m, l H ) , 8.17 in order to measure inhibitory a c t i v i t i e ~ .IC50 ~ ~ values were (dt, l H , J = 7.6, l.O), 7.88-7.76 (m, 2H), 7.75 (qd, l H , J = calculated from a hyperbolic equation described elsewhere.24 7.6, 1.1,l.O), 7.65 (td, l H , J = 7.6, 1.21, 7.45 (td, l H , J = 7.6, Time dependence was tested by preincubating for 5 and 15 1.0); IR (cm-') 1720, 1390. Anal. (C15H~N20)C, H, N. 3-(2'-Methoxyphenyl)-5H-indeno[ 1,2-clpyridazin-5- min a t 37 "C and a t a single concentration of the inhibitor. Reversibility was demonstrated by measuring the degree of one (32). 2'-Methoxyacetophenone (300 mg, 2 mmol) and 356 inhibition in the presence of a higher substrate concentration mg (2 mmol) of ninhydrin were refluxed in acetic acid ( 4 mL) (540 u M ) . for 2 h. The solution was cooled, and the solid obtained was Incubations were carried out a t pH = 7.4 (Na,HPOdKH2collected and washed with diethyl ether. This crude aldol Po4 isotonized with KCl) a t 37 "C. The mitochondrial suspenadduct was dissolved in THF (40 mL), and 0.1 mL (2 mmol) of sion was set to a final protein concentration of 1.0 mg/mL. The hydrazine monohydrate was added under magnetic stirring. mitochondria were preincubated a t 37 "C for 5 min with After 2 h the mixture was evaporated under vacuum and the clorgyline (250 nM) o r selegeline (250 nM) to test MAO-B or residue was purified by flash chromatography on silica gel MAO-A activity, respectively. The inhibitor under study was using CHC13 a s eluent: yield 24%; mp 162-3 "C, from CHCld solubilized in DMSO, added to give a final DMSO concentrahexane; 'H-NMR (CDC13)6 8.16 (dt, l H , J = 7.6, 0.91, 8.14 ( s , tion of 5% (v/v), and further incubated for 5 min. Preliminary l H ) , 7.96 (dd, l H , J = 7.6, 1.81, 7.83 (qd, l H , J = 7.6, 1.2, experiments had verified that DMSO a t this concentration 0.9), 7.71 (td, 1 H , J = 7.6, 1.21, 7.53 (td, 1 H , J = 7.6, 0.91, 7.46 does not affect MA0 activity. The nonselective substrate (qd,lH,J=8.3,7.6,1.8),7.11(td,lH,J=7.6,1.0),7.02(dd, kynuramine is deaminated by MA0 to an aldehyde that l H , J = 8.3, J = 1.0),3.88( s , 3H); IR (cm-') 1730, 1410. Anal. spontaneously cyclizes t o 4-hydroxyquinoline. Formation of (ClsH12N202) C, H, N. 3-(3'-Aminophenyl)-5H-indeno[ 1,2-c]pyridazin-5-one the latter was monitored continuously a t 314 nm using a Kontron UVIKON 941 spectrophotometer (Kontron AG, Basel, (37). A solution of crude 3-[3'-(N-carbobenzoxyamino)phenyllSwitzerland). Incubations were then carried out with a 5H-indeno[l,2-clpyridazin-5-one (814 mg, 2 mmol), prepared substrate concentration of KM (KM= 90 ,uM for MAO-A, 60 pM according to method B from N-carbobenzoxy-3'-aminoacetophefor MAO-B) without inhibitor or with different inhibitor none, was refluxed over 1.5 h in 20 mL of aqueous HBr/acetic concentrations. acid (50:50). The mixture was cooled a t room temperature and QSAR Study. Multilinear regression and cluster analysis made basic with KOH, and the red solid obtained was were carried out with TSAR 2.1 software (Oxford Molecular, separated by filtration and purified by flash chromatography
3D-QSARsof MAO-B Inhibition Oxford, U.K.), and PLS analysis was performed using the Q S A R module in SYBYL 6.04 software (Tripos Assoc., St. Louis, MO). Both program were run on Silicon Graphics Personal Iris 4D-35 and Silicon Graphics Indigo R4000 workstations. The u values were collected from standard compilat i o n ~ If . ~not~ available, ~ ~ ~ the F , R, and u parameters for ortho substituents were calculated according t o Williams and Norrington.28 CoMFA (3D-QSAR) Study. Molecular models of the IP analogs listed in Tables 2 and 3 were constructed using SYBYL 6.04 software (Tripos Assoc., St. Louis, MO) running on Silicon Graphics Personal Iris 4D-35 and Silicon Graphics Indigo R4000 workstations. The geometry of each molecule was minimized using the standard Tripos force field29 including the electrostatic energy term calculated with the GasteigerMarsili atomic charges.30 From a conformational analysis performed with the systematic search option, the lowest energy minimum was selected for each compound and minimized. The CoMFA studies were carried out by using the QSAR module of SYBYL version 6.04. Default settingsz0were used in the analyses, except for the option ‘‘drop-electrostaticS’’ which was set t o “no”. The grid size had a resolution of 1.5 A. As a third field, the Molecular Lipophilicity Potential (MLP) was used.31 Only those compounds for which an IC50 value could be determined (i.e., 39 compounds in Tables 2 and 3) were used in the analyses. Only models with q2 > 0.4 were considered as significant. To avoid reducing the training set, we took as test set the two compounds (16 and 56) which had a degree of inhibition larger than 25% a t their maximum solubility, thus allowing their IC50 values to be calculated with reasonable confidence by hyperbolic extrapolation.
Acknowledgment. B.T. and P.A.C. are grateful to the Swiss National Science Foundation for support. The Italian authors gratefully acknowledge support from MURST and CNR (Rome, Italy). References (1) Sandler, M. Monoamine oxidase inhibitor efficacy in depression and the “cheese effect”. Psychol. Med. 1981,11, 455-458. (2) Kyburz, E. New developments in the field of MA0 inhibitors. Drug News Perspect. 1990,3,592-599. (3) Youdim, M. B. H.; Finberg, J. P. M. New directions in monoamine oxidase A and B selective inhibitors and substrates. Biochem. Pharmacol. 1991,41,155-162. (4) Cesura, A. M.; Pletscher, A. The new generation of monoamine oxidase inhibitors. In Progress in Drug Research, Vol. 38; Jucker, E., Ed.; Birkhauser Verlag: Basel, Switzerland, 1992; pp 171-297. (5) DaPrada, M.; Kettler, R.; Haefely, W. E. Some basic aspects of reversible inhibitors of monoamine oxidase-A. Acta. Psychiatr. Scand. 1990,360 (Suppl.), 7-12. (6) Haefely, W. E.; Burkard, W. P.; Cesura, A. M.; Kettler, R.; Lorez, H. P.; Martin, J. R.; Richards, J. G.; Scherschlicht, R.; DaPrada, M. Biochemistry and pharmacology of moclobemide, a prototype RIMA. Psychopharmacology 1992,106,6-14. (7) McDonald, I. A.; Bey, P.; Palfreyman, M. G. Monoamine oxidase inhibitors. In Design of Enzyme Inhibitors as Drugs; Sandler, M., Smith, H. J., Eds.; Oxford University Press: Oxford, U.K., 1989; pp 227-243. (8) Tipton, K. F.; Fowler, C. J. The kinetics of monoamine oxidase inhibitors in relation to their clinical behaviour. In Monoamine Oxidase and Disease; Prospects for Therapy with Reversible Inhibitors; Tipton, K. F., Dostert, P., Strolin Benedetti, M., Eds.; Academic Press: London, U.K., 1984; pp 27-40. (9) Heinisch, G.; Kopelent-Frank, H. Pharmacologically active pyridazine derivatives. In Progress in Medicinal Chemistry, Vol. 29;Ellis, G.P., Luscombe, D. K., Eds.; Elsevier: Amsterdam, The Netherlands, 1992; pp 141-167. (10) Mazouz, F.; Lebreton, L.; Milcent, R.; Burstein, C. Inhibition of monoamine oxidase types A and B by 2-aryl-4H-1,3,4-oxadiazin5(6H)-onederivatives. Eur. J . Med. Chem. 1988,23,441-451.
Journal of Medicinal Chemistry, 1995, Vol. 38, No. 19 3883 (11) Carotti, A,; Carta, A,; Campagna, F.; Altomare, C.; Casini, G.
An efficient route to biologically active 5H-indeno[l,2-cl-pyridazin-5-ones. Farmaco 1993,48, 137-141. (12) Carotti, A.; Altomare, C.; Campagna, F.; Carta, V.; Kneubuhler, S.; Thull, U. Synthesis and monoamine oxidase inhibitory activity of 5H-indeno[l,2-clpyridazines.Abstracts of the 12th International Symposium on Medicinal Chemistry;Swiss Chemical Society: Basel, Switzerland, 1992; p 304. (13) Dixon, C. M.; Park, G. R.; Tarbit, M. H. Characterization of the enzyme responsible for the metabolism of sumitriptan in human liver. Biochem. Pharmacol. 1994,47,1253-1257. (14) Mazouz, F.; Lebreton, L.; Milcent, R.; Burstein, C. 5-Aryl-1,3,4oxadiazol-Z(3H)-one derivatives and sulfur analogues as new selective and competitive monoamine oxidase type B inhibitors. Eur. J . Med. Chem. 1990,25,659-671. (15) Mazouz, F.; Gueddari, S.; Burstein, C.; Mansuy, D.; Milcent, R. 5-(4-Benzyloxy)pheny1)-1,3,4-oxadiazol-2(3H)-one derivatives and related analogues: new reversible, highly potent, and selective monoamine oxidase type B inhibitors. J . Med. Chem. 1993,36, 1157-1167. (16) Thull, U.; Testa, B. Screening of unsubstituted cyclic compounds as inhibitors of monoamine oxidases. Biochem. Pharmacol. 1994,47, 2307-2310. (17) Kneubuhler, S.; Carta, V.; Altomare, C.; Carotti, A.; Testa, B. Synthesis and monoamine oxidase inhibitory activity of 3-substituted 5H-indeno[l,2-clpyridazines.Helv. Chim. Acta 1993, 76, 1954-1963. (18) Van de Waterbeemd, H.; Testa, B. The parametrization of lipophilicity and other structural properties in drug design. In Advances in Drug Research, Vol. 16;Testa, B., Ed.; Academic Press: London, U.K., 1987; pp 85-225. (19) Hansch, C.; Leo, A. Substituent Constants for Correlation Analysis i n Chemistry and Biology; Wiley: New York, 1979. (20) Cramer, R. D.; Patterson, D. E.; Bunce, J. D. Comparative molecular field analysis (CoMFA). 1. Effect of shape on binding of steroids to carrier proteins. J . Am. Chem. SOC.1988,110, 5959-5967. (21) DaPrada, M.; Kettler, R.; Keller, H. H.; Cesura, A. M.; Richards, J. G.; Saura Marti, J.;Muggli-Maniglio, D.; Wyss, P. C.; Kyburz, E.; Imhof, R. From moclobemide to Ro 19-6327 and Ro 41-1049: the development of a new class of reversible, selective MAO-A and MAO-B inhibitors. J . Neural. Transm., Suppl. 1990,29, 279-292. (22) Clark, J. B.; Nicklas, W. J. The metabolism of rat brain mitochondria. J . B i d . Chem. 1970,245,4724-4731. (23) Walther, B.; Ghersi-Egea, J.; Minn, A,; Siest, G. Brain mitochondrial cytochrome P-45Oscc: Spectral and catalytic properties. Arch. Biochem. Biophys. 1987,254,592-596. (24) Thull, U.; Kneubuhler, S.; Testa, B.; Borges, M. F. M.; Pinto, M. M. M. Substituted xanthones as selective and reversible monoamine oxidase A (MAO-A) inhibitors. Pharm. Res. 1993, 10,1187-1190. (25) Lowry, 0.;Rosebrough, N.; Lewis Farr, A.; Randall, R. Protein measurement with the folin uhenol reagent. J. Biol. Chem. 1951,193,265-275. (26) Weissbach. H.; Smith, T. E.; Daly, J. W.; Witkop, B.; Udenfriend, S. A rapid spectrophotometric assay of monoamine oxidase based on the rate of disappearance of kynuramine. J . B i d . Chem. 1960,235,1160-1163. (27) Perrin, D. D.; Dempsey, B.; Serjeant, E. P. p K , Prediction for Organic Acids and Bases; Chapman and Hall: New York, 1981. (28) Williams, S. G.; Norrington, F. E. Determination of positional weighting factors for the Swain and Lupton substituent constants F and R. J . Am. Chem. SOC. 1976,98,508-516. (29) Clark, M.; Cramer, R. D.; Van Opdenbosch, N. Validation of the general purpose Tripos 5.2 force field. J . Comput. Chem. 1989, 10,982-1012. (30) Gasteiger,J.; Marsili, M. Iterative partial equalization of orbital electronegativity: a rapid access to atomic charges. Tetrahedron 1980,36,3219-3222. (31) Gaillard, P.; Carrupt, P.-A.; Testa, B.; Boudon, A. Molecular lipophilicity potential, a tool in 3D QSAR: Method and applications. J . Comput.-Aided Mol. Des. 1994,8,83-96. I
JM950172K
